PROJECT SUMMARY/ABSTRACT Many persons who die by suicide access health care services in the year before the death, indicating potential opportunity for improved detection and intervention in health care settings. Research is needed to fill gaps in our understanding of how precursors and health care services use patterns may help to inform identification of those at highest risk. This proposal focuses on determining precursors and health care services use patterns, and moderators of these patterns, following release from prison and transition into community living - a circumstance associated with exceptionally high rates of suicide and death by unintentional injury, including drug overdose. We propose an innovative method of linking administrative health care use datasets from Medicare and the VA (2008 ? 2016) to identify a ?prison release? cohort whose health care utilization (and that of a control cohort matched on age, sex, race, residential zip code, and Medicare status) can be tracked and compared over time. We will strategically target individuals who are age 50 and older and are dually eligible for Medicare Fee-for-Service (FFS) and Veterans Health Administration (VHA) health services because: older released prisoners are largely older reentry veterans; among both veterans and the general population, suicide rates are highest among those 50 and older; prison release confers high risk of suicide and death by unintentional injury; and we can evaluate those eligible for Medicare FFS based on age (65 years and older) or based on disability. Then, capitalizing on the availability of suicide surveillance data routinely collected by the VA, namely the National Suicide Data Repository (SDR) and the Suicide Prevention Applications Network (SPAN), we can rigorously evaluate suicide, death by unintentional injury (distinguishing death by drug overdose), and suicide attempt. Together, the pooled data from multiple national-level sources will enable us to comprehensively examine precursors (i.e., predisposing, enabling, and need factors) and distinguish health care use patterns that may increase or mitigate risk of experiencing suicide-related outcomes in a highly vulnerable population. Our analytic approach will incorporate innovative longitudinal modeling techniques, including prognostic modeling in the context of survival analysis methods that account for competing risk of death from other causes. Furthermore, to assess near- and longer-term risk following prison release, precursors and patterns of health care use associated with risk of suicide-related outcomes will be examined over the entire study period, as well as at different salient time intervals (e.g., within 90-days, 1-year). The potential public health significance of this study is substantial. Findings from the proposed research can support suicide prevention efforts by informing prison-to-community transition care planning, with implications for multiple stakeholders. Furthermore, improved understanding of the rate and nature of suicide events within and between the two largest national health care systems will provide metrics that can be used to inform prevention and intervention efforts to drive down suicide events in these systems.